1
MANAGING POPULATIONS WITH AN
AUTOMATED CLINICAL RISK SYSTEM
Session #304, February 15, 2019
John Dunlap, DO, CPHQ, CCHP, Deputy Medical Executive, California Correctional Health Care Services
Michael Selby, MBA, Health Program Manager III, California Correctional Health Care Services
2
John Dunlap, DO, CPHQ, CCHP
I have no conflicts of interest to report. I have no financial
interest in any device, software, corporation, or research firm.
Michael Selby, MBA
I have no conflicts of interest to report. I have no financial
interest in any device, software, corporation, or research firm.
CONFLICT OF INTEREST
3
• Identify the healthcare factors utilized by CCHCS to stratify clinical
risk
• Translate the change management strategy used by CCHCS to
implement the Automated Clinical Risk Classification System for
use in your organization’s informatics projects
• Recognize the benefits of co-locating clinical subject matter
experts with programmers
LEARNING OBJECTIVES
4
A G E N D A
• LOCAL PROBLEM
• DESIGN & IMPLEMENTATION
• HOW IT SOLUTION WAS UTILIZED
• VALUE DERIVED
5
L O C A L P R O B L E M
6
P R O B L E M S T A T E M E N T
High risk population are not consistently and accurately identified and placed throughout
their incarceration to receive the right level of resources.
I
B B
B
B
B
B
B
B
B B
B
B
B B
B
B
B
B B
B
B
B
B
I
I
I
I
I
I
I
I
I
I
I
7
E V O L U T I O N O F M E D I C A L C L A S S I F I C A T I O N
2006
Receivership
Began
2009
Medical
Classification
System Introduced
2012
Medical
Classification
e-Form Introduced
2013
Automated Clinical
Risk Classification
Implemented
Healthcare Factors Introduced
• Rules Broadly Defined Risk & Open To
Interpretation
• Often Inaccurate
• Often Not Up To Date
• Person Dependent Paper Process in
Communicating to Custody
Little to no
consideration for
Healthcare Factors
Communication to Custody Improved
• Paper Process Eliminated, But Still
Person Dependent
• Rules Still Not Standardly Applied
• Inaccuracies Persisted
• Still Not Updated Timely
8
Provider conducts
chart review to
determine clinical risk.
Provider manually enters
clinical risk into Medical
Classification e-Form.
e-Form transmits information
to SOMS; custody uses
healthcare factors to
appropriately house inmate.
When clinical status
changed: Provider
manually updates e-Form.
P R I O R T O N E W S O L U T I O N . . .
Repeated audits found discrepancies between reported risk and actual risk.
9
• Only 62% of High Risk Patients were Properly Placed
• 6/2013 Hospital Readmission Rate
20% High Risk vs. 14% Total Population
• 6/2015 Potentially Avoidable Hospitalizations (per 1,000)
71 High Risk vs. 11 Total Population
B A S E L I N E D A T A
10
IMPROVE CARE & REDUCE COSTS
CLINICAL
• Increase Appropriate Placement of High
Risk Inmates
• Reduce Preventable Morbidity & Mortality
LEVERAGE IT & DATA
ADMINISTRATIVE
• Automate Complicated Rules for Clinical
Risk
• Provide Timely, Actionable Information
• Provide Near Real-Time Performance
Reports
E X P E C T E D P R O J E C T O U T C O M E S
11
D E S I G N &
I M P L E M E N T A T I O N
12
E V A L U A T I N G I T S O L U T I O N A L T E R N A T I V E S S
COTS SOLUTIONS WERE NOT VIABLE
• Risk stratification applications based on Claims Data from PCPs
– CCHCS PCPs Don’t Bill for Services
• Unique requirements for correctional setting
EXISTING RESOURCES
• E-Form Application
• Healthcare Data Warehouse – SQL
• In-House Talent (Clinical and Technical)
13
D E V E L O P I N G T H E I T W O R K F L O W
Patient
Demographics
Q 12 HOURS
HEALTHCARE DATAWAREHOUSE
Automated job runs to refresh data tables
CUSTODY
Notified of Clinical Risk
changes automatically
SOMS
eFORM
APPLICATION
Health Records
Claims
Diagnostics
14
DEFINE RULES FOR EACH FACTOR
DEFINE HIGH RISK FACTORS
DEFINE RISK LEVELS & AVAILABLE SOURCES
DEFINE DATA ELEMENTS & SOURCES FOR FACTORS
• Medium Risk Outpatient
• Low Risk Outpatient
• High Risk + Inpatient/Specialized Care Needs
• High Risk Outpatient
• Chronic Medical Conditions & Degree of Control
• Hospitalizations/Emergency Dept. Encounters
• 4 or more conditions; >1 problem prone condition
• Advanced Age- >65
• Problem List Entries
• Custody Demographic Information
• Age
• Medications
• Polypharmacy (12+ Medications)
• 2+ Hospital/3+ ED Encounters
• Pharmacy and Diagnostic Information Systems
• Claims Data
H O W U S E R S S U P P O R T E D S O L U T I O N
& W O R K F L O W D E V E L O P M E N T
15
H O W U S E R S S U P P O R T E D S O L U T I O N
& W O R K F L O W D E V E L O P M E N T
16
H O W U S E R S S U P P O R T E D S O L U T I O N
& W O R K F L O W D E V E L O P M E N T
17
PARTNERING BUSINESS OWNER WITH IT DEVELOPERS
PROGRAMMERS
CLINICAL SMEs
18
H O W I T S O L U T I O N
W A S U T I L I Z E D
19
I T S O L U T I O N F U N C T I O N A L I T Y
20
I T S O L U T I O N F U N C T I O N A L I T Y C O N T .
21
I T S O L U T I O N F U N C T I O N A L I T Y C O N T .
22
C L I N I C A L W O R K F L O W I M P A C T E D
Provider conducts
chart review to
determine clinical
risk.
Provider manually enters
clinical risk into Medical
Classification e-Form.
e-Form transmits
information to SOMS;
custody uses healthcare
factors to appropriately
house inmate.
When clinical status changed:
Provider manually updates e-
Form.
23
I T S O L U T I O N W I T H I N T H E W O R K F L O W
Custody Staff
reviews Custody
Registry
Correctional
Counselor reviews
patient’s record
Correctional Counselor
determines appropriate
housing
Custody
moves patient
Risk Level
Automatically
Updated
24
I T S O L U T I O N B L U E P R I N T
DIAGNOSES
LAB DATA
PHARMACY
DATA
PATIENT
MEDICAL
CONDITIONS
PATIENT
OVERALL
RISK LEVEL
LOW
HIGH 2
MEDIUM
HIGH 1
HEALTHCARE
TOOLS
DASHBOARD
MEASURE
CUSTODY
REGISTRY
CREATE INFRASTRUCTURE
DESIGN OUTPUT
DEVELOP TOOLS
PATIENT
DEMOGRAPHICS
25
C C H C S H E A L T H C A R E D A T A W A R E H O U S E
DATA
WAREHOUSE
• Metadata
• Raw Data
• Summary Data
AD HOC ANALYSIS
REPORTING
DATA MINING
EHRS
SOMS
LEGACY
SYSTEMS
FLAT FILES
ETL
Extract
Transform
Load
26
C H A N G E M A N A G E M E N T S T R A T E G Y
• Organizational memorandum from leadership
• Executive forums at all levels of the organization to communicate this change
• Refine existing policy and procedures
• Define process workflows
• Override process
• Decision support
• Training materials
• Status reports post implementation to see progress upon initial implementation
• Statewide high risk performance report
27
C H A N G E M A N A G E M E N T S T R A T E G Y
PROVIDING ACCESS TO END-USERS – LINKS FROM CDCR INTRANET TO QM TOOLS
28
V A L U E D E R I V E D
29
P O S T I M P L E M E N T A T I O N P R O C E S S
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
6/1/13
9/1/13
12/1/13
3/1/14
6/1/14
9/1/14
12/1/14
3/1/15
6/1/15
9/1/15
12/1/15
3/1/16
6/1/16
9/1/16
12/1/16
3/1/17
6/1/17
9/1/17
12/1/17
3/1/18
6/1/18
18%
High Risk
Placement
30
I M P R O V E M E N T S A C H I E V E D
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
6/1/13
9/1/13
12/1/13
3/1/14
6/1/14
9/1/14
12/1/14
3/1/15
6/1/15
9/1/15
12/1/15
3/1/16
6/1/16
9/1/16
12/1/16
3/1/17
6/1/17
9/1/17
12/1/17
3/1/18
6/1/18
High Risk Rules
Modified and
Conversion to ICD-10
31
P O S T I M P L E M E N T A T I O N O U T C O M E
0%
5%
10%
15%
20%
25%
Statewide Hospital Readmissions
June 2013 – December 2017
All Hospital Readmissions High Risk Hospital Readmissions
All Hospital Readmissions High Risk Hospital Readmissions
June 2013 14% 20%
December 2017
12% 16%
32
P O S T I M P L E M E N T A T I O N O U T C O M E
0
10
20
30
40
50
60
70
80
Statewide Potentially Avoidable Hospitalizations (per 1,000)
June 2015 – December 2017
All PAH High Risk PAH
All Potentially Avoidable
Hospitalizations
High Risk Potentially Avoidable
Hospitalizations
June 2015 10.8 70.9
December 2017
8.9 43.5
33
E S T I M A T E D C O S T
FIXED COSTS
IT
Costs
(
Storage, Software)
$525,000/year
DBA/DBD Contract
$125,000/year
Total
$650,000/year
DEVELOPMENT LABOR COSTS
QM Programmer
$90,000
Clinical SME
$62,500
Total
$152,500
Note: Fixed costs are associated to HDW & all applications; not specific to this solution
M&O LABOR COSTS
QM
Programmer/year
$10,000
Clinical SME/year
$20,000
Total
$30,000
34
C O S T S A V I N G S
ESTIMATED COSTS SAVED BY AUTOMATING RISK LEVEL
– 230 Clinical Risk Changes per day x 5 minutes
clinician would spend reviewing patient’s chart
– 12 e-Forms per hour x $100 per hour (for P&S
classification)
– $1900 in saved time per day
– $2.1 million saved over 3 years, and growing
35
E X T E R N A L S T U D I E S C O M P A R I S O N S
Many models use the following variables:
 Age
 Gender
 Drugs
 Diagnoses
Some also use:
 Disability/functional status
 Procedures
Insurance type
Region
Employment info
Cost/use data
Not many (or none) use the following that we include:
 Lab values
 Provider specialty
None use as many factors as we do (for better or worse)
36
U S E R F E E D B A C K
37
L E S S O N S L E A R N E D
• Data is an asset that should be managed and leveraged
• Risk Classification should not be static
o Need process for resolving disputes
o Rules should be reevaluated periodically
• With the right IT infrastructure a healthcare organization can start to do
large scale improvement efforts
• Timeframe for goal overly ambitious
o Competing placement factors
o Patient population & individual clinical risk change over time
• Change management strategy may need to be revisited
o Focus and engage with custody stakeholders more
38
Q U E S T I O N S ?